Abstract
Diplegia is a specific subcategory of the wide spectrum of motion disorders gathered under the name of cerebral palsy. Recent works proposed to use gait analysis for diplegia classification paving the way for automated analysis. A clinically established gait-based classification system divides diplegic patients into 4 main forms, each one associated with a peculiar walking pattern. In this work, we apply two different deep learning techniques, namely, multilayer perceptron and recurrent neural networks, to automatically classify children into the 4 clinical forms. For the analysis, we used a dataset comprising gait data of 174 patients collected by means of an optoelectronic system. The measurements describing walking patterns have been processed to extract 27 angular parameters and then used to train both kinds of neural networks. Classification results are comparable with those provided by experts in 3 out of 4 forms.
Highlights
Cerebral palsy (CP) is a group of permanent movement disorders appearing in early childhood
A specific type of artificial neural network, known a selforganizing map (SOM), has been exploited in [11] to learn an abstract representation of a normal gait pattern from a dataset of 129 gait cycles acquired from 18 typically developed subjects. e method developed for classification is based on the idea that the degree of motor disorder can be identified by evaluating the quantisation error (QE) of the differences between normal and abnormal gait patterns
Experimental Results e performance of the proposed multilayer perceptron (MLP) network and recurrent neural networks (RNNs) were assessed on a Nvidia GTX-1060 board using Keras [22]
Summary
Cerebral palsy (CP) is a group of permanent movement disorders appearing in early childhood. Top-down components are intended as the main features of motor organization spontaneously performed by individuals in order to find and maintain a safe, stable, and efficient gait These are (i) support reaction to body weight, i.e., pattern adopted to find and maintain balance, (ii) static and dynamic fixation mechanisms, i.e., modality of use of canes, and position and swing of upper limbs, (iii) step-bystep righting reactions, i.e., trunk and pelvis pendulum and translation in and out of sagittal plane, (iv) mechanisms of progression, i.e., choice of fulcra upon which to pivot the body, and (v) ability to correctly integrate proprioceptive and perceptive information [3, 5].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
